This 2017 review surveys 17 trials that tested the effects of a very low calorie diet (VLCD) on people with diabetes. A VLCD is defined as a diet of fewer than 800 calories per day.
These 17 trials, published from 1986 to 2012, vary widely in design; daily calorie intake ranged from as low as 300 calories to as high as 800 calories, and trials ranged from as few as eight days to as many as six months. As such, summary statistics (i.e., the mean impact of a VLCD across all studies on specific biomarkers) were not calculated.
Seven of 14 trials that looked at fasting glucose showed that a VLCD led to significant improvements, while three of four found the same for fasting insulin. Nine studies looked at blood pressure, and all showed significant improvements. Ten of 14 studies that tracked blood lipids showed significant reductions in total cholesterol and triglyceride levels. No adverse effects linked to VLCDs were reported, though dropout rates ranged from 4.7% to 33%. (Note: Higher dropout rates were seen in longer trials.)
This review shows the difficulty of using existing evidence to clearly assess the impact of a VLCD on diabetes, in large part because the parameters of a VLCD have not been consistently defined. However, the review also shows that a number of trials using VLCDs demonstrated rapid, clinically meaningful improvements in metabolic markers that were sustained even after the VLCD concluded.
For example, four trials showed large improvements in fasting glucose and HbA1c due to a VLCD. In brief, these trials are:
- Henry et al., 1986: A 300-calorie/d diet for 36 days decreased fasting glucose from 16.2 mmol/L to 5.3 mmol/L and HbA1c from 13.1% to 8.8%;
- Wing et al., 1991: A 400-calorie/d diet for eight weeks decreased fasting glucose from 14.2 mmol/L to 7.7 mmol/L and HbA1c from 10.4% to 7.3%;
- Kelley et al., 1993: A 400- to 800-calorie/d diet for three months decreased fasting glucose from 223 mg/dL to 109 mg/dL and HbA1c from 8.8% to 6.3%;
- Snel et al., 2011: A 450-kcal/d diet for four months decreased fasting glucose from 12.1 mmol/L to 7.7 mmol/L and HbA1c from 7.8% to 6.7%. When this diet was combined with exercise, fasting glucose fell to 6.6 mmol/L and HbA1c to 6.3%.
Similar interventions may provide safe, rapid improvements in diabetes, particularly when paired with proper protein dosing and resistance training to prevent muscle loss.
A Systematic Review of Evidence on the Use of Very Low Calorie Diets in People with Diabetes